mlx-community/Qwen3-Coder-30B-A3B-Instruct-8bit
Quantization vs Qwen/Qwen3-Coder-30B-A3B-Instruct. mlx-community/Qwen3-Coder-30B-A3B-Instruct-8bit is a quantized version of the other (U32/BF16).
3 other repos hold byte-for-byte identical weight files.